Professional Documents
Culture Documents
LECTURE 32
1. Back to Systems
We return our attention now to the system of equations
a11 x1 + a12 x2 + . . . + a1n xn = b1 (32.1)
a21 x1 + a22 x2 + . . . + a2n xn = b2
..
.
an1 x1 + an2 x2 + . . . + ann xn = bn
To express this system of equations in matrix form, we start by writing both sides as vectors.
a11 x1 + a12 x2 + . . . + a1n xn b1
a21 x1 + a22 x2 + . . . + a2n xn b2
= ..
..
. .
an1 x1 + an2 x2 + . . . + ann xn bn
Notice that the left hand side of the equation can be rewritten as a matrix-vector product.
a11 a12 . . . a1n x1 b1
a21 a22 . . . a2n x2 b2
.. .. = ..
.. .. ..
. . . . . .
an1 an2 . . . ann xn bn
We can simplify this notation by writing
Ax = b (32.2)
where x is the vector whose entries are the variables in the system, A is the matrix of coefficients
of the system (called the coefficient matrix ), and b is the vector whose entries are the right-hand
sides of the equations. We call (32.2) the matrix form of the system of equations (32.1).
We know that the system of equations (32.1), and hence (32.2), have zero, one, or infinitely
many solutions. Suppose det(A) 6= 0, i.e., A is nonsingular. Then (32.2) has only one solution,
namely,
x = A−1 b.
So we can rewrite our earlier Theorem 31.1 in the following way.
Theorem 32.1. Given the system of equations (32.2),
(1) if det(A) 6= 0, there is exactly one solution;
(2) if det(A) = 0, there are either zero or infinitely many solutions.
1
Differential Equations Lecture 32: More Linear Algebra
Recall that if our system (32.1) were homogeneous, i.e., if each bi = 0, we always have the trivial
solution xi = 0. Denoting the vector with entries all 0 by 0, the matrix form of a homogeneous
system is
Ax = 0. (32.3)
Thus we can express the earlier Theorem 31.2 as follows.
Theorem 32.2. Given the homogeneous system of equations (32.3),
6 0, there is exactly one solution, x = 0;
(1) if det(A) =
(2) if det(A) = 0, there will be infinitely many nonzero solutions.
2
Differential Equations Lecture 32: More Linear Algebra
Next, we’ll write out the components of the two vectors and multiply through.
−2 4 η1 0
=
2 −4 η2 0
−2η1 + 4η2 0
=
2η1 − 4η2 0
3
Differential Equations Lecture 32: More Linear Algebra
For this vector equation to hold, the components must match up. So we’ve got to find a
solution to the system
−2η1 + 4η2 = 0
2η1 − 4η2 = 0.
Notice that these are the same equation: the second line only differs from the first by
multiplication by a constant, in this case −1. This will always be the case if we’ve found
our eigenvalues correctly, since we know that A − λI is singular and so our system (32.5)
should have infinitely many solutions. In fact, going back to our matrix A − λI, we could
notice there that the rows only differed by a constant factor. This is a good place to check
that our earlier algebra was correct: if the rows differ by more than just a constant factor,
something’s gone wrong.
Since these equations are identical, we can just choose one (whichever is convenient
works fine) and obtain a relation between the eigenvector components η1 and η2 . Let’s
choose the first. This gives
2η1 = 4η2 ,
and so we have η1 = 2η2 . As a result, any eigenvector corresponding to λ1 = 5 has the
form
η1 2η2
η= = .
η2 η2
There are infinitely many vectors of this form, of course; we only need one. We can select
one by choosing a value for η2 . The only restriction is that we shouldn’t take η2 = 0, since
then η = 0, which we don’t want. We may choose, for example, η2 = 1, and then we have
(1) 2
η = .
1
(2) λ2 = −1 :
In the previous case, we went into more detail than we will in future examples. The
process is exactly the same, however.
Plugging in λ2 into (32.5) gives the system
4 4 η1 0
=
2 2 η2 0
4η1 + 4η2 0
= .
2η1 + 2η2 0
The two equations corresponding to this vector equation are
4η1 + 4η2 = 0
2η1 + 2η2 = 0.
Once again, these only differ by a constant factor: the first equation is twice the second.
Let’s choose the second equation to work with, though in this case it doesn’t matter at
all; both are equally easy. We have
η1 = −η2 ,
and so any eigenvector has the form
η1 −η2
η= = .
η2 η2
4
Differential Equations Lecture 32: More Linear Algebra
5
Differential Equations Lecture 32: More Linear Algebra